EXCEED THE SPACE PROVIDED. This is a revised, competing renewal for a project funded in 1994 (R01 MH50038; MRI supplement funded in 1995). Mild Cognitive Impairment (MCI) is a common clinical problem for which there are limited data on predictors of who will develop Alzheimer's Disease (AD) on follow-up. In a sample of 150 MCI outpatients and 75 matched normal controls followed for 4-9 years, the following potential baseline predictors of AD are evaluated: known demographic (e.g., age) and clinical (e.g., head injury) risk factors, impairment on specific neuropsychological tests, olfactory detection deficits, hippocampal atrophy on MRI scan of the brain, and temporoparietal flow deficit on resting SPECT scan of the brain. The relative accuracy (sensitivity and specificity) of these predictors will be determined regarding which MCI patients meet diagnostic criteria for AD during follow-up. In addition, for clinical use, a predictive algorithm for AD will be developed based on a combination of these potential early diagnostic markers. The relative utility of these predictors in clinical diagnostic decision-making will also be evaluated using receiver operating characteristic (ROC) curves, e.g., how much does each variable add to predictive accuracy relative to that obtained with a screening instrument like the Mini Mental State Exam, plus or minus neuropsychological testing? To address these aims, we have recruited and followed 94 MCI patients and 46 normal controls. We will increase the sample size to 150 MCI patients and 75 normal controls, and continue systematic follow-up to ensure that nearly all possible conversions to AD (n=19 to date) are identified. We estimate that at the end of 4-9 years' follow-up, approximately 50% of the MCI sample will meet diagnostic criteria for AD and 40% will not meet criteria for AD (5% dropout and 5% other dementias), resulting in an excellent distribution to test the hypotheses. In the ongoing study, each of the baseline predictors has already predicted AD, or Clinical Dementia Rating decline, during follow-up. A 4-year follow-up MRI and SPECT (with neuropsychological testing) is added to assess progression of hippocampal atrophy and temporoparietal blood flow deficit, which will provide support for the diagnosis of AD in patients who are not deceased and hence do not reach autopsy. This study will lead to improved clinical accuracy for the early diagnosis of AD, and will eventually help enrich high-risk patient selection for clinical trials of therapeutic agents in MCI.